PURPOSE Cancer patients are at an increased risk of venous thromboembolism (VTE). VEGF promotes the growth of highly thrombogenic tumor vessels. Here, we explored the utility of soluble plasma VEGF-A (sVEGF) as a biomarker for the prediction of VTE in patients with cancer. EXPERIMENTAL DESIGN Eight hundred four patients with newly diagnosed cancer or progression after remission were prospectively followed for 2 years until the occurrence of VTE or death [tumor sites: brain (n = 87), breast (n = 137), lung (n = 120), gastrointestinal (n = 143), pancreas (n = 53), prostate (n = 95), kidney (n = 22), myeloma and lymphoma (n = 99), and others (n = 48)]. Primary endpoint was symptomatic or fatal VTE. sVEGF was measured by immunoassay in baseline plasma. RESULTS Fifty-five patients developed VTE (6.8%) and 364 patients (45.3%) died. Five-hundred and forty-two (68.3%) participants had sVEGF levels above the detection limit of 0.5 pg/mL. The median sVEGF level (25th-75th percentile) was 8.1 pg/mL (0-17.7). The cumulative 2-year incidence of VTE was 10.2% [95% confidence interval (CI), 6.4-14.9] in patients with sVEGF greater than the 75th percentile of the sVEGF distribution (Q3, cutoff: 17.7 pg/mL), and 5.9% (95% CI, 4.2-7.9) in patients with lower levels (P = 0.03). The corresponding 2-year risk of death was 52.8% (95% CI, 46.0-60.0) and 43.9% (95% CI, 40.0-48.0), respectively (P = 0.02). In univariable time-to-VTE regression, elevated sVEGF was associated with VTE [subhazard ratio (SHR) per 10 pg/mL increase, 1.04; 95% CI,1.00-1.09; P = 0.04)]. The association between sVEGF and risk of VTE prevailed after multivariable adjustment for high-risk tumor sites, age, gender, factor VIII, thrombin generation potential, and soluble P-Selectin (adjusted SHR, 1.04; 95% CI, 1.00-1.09, P = 0.05). CONCLUSIONS Elevated sVEGF is associated with an increased risk of VTE in patients with cancer.